A Methodological Map of Personalized Cancer Vaccines

A new essay on cancer vaccines explains the path from tumor mutations to personalized therapy

On September 2, Abhishaike Mahajan, a machine learning engineer at a biotechnology company, published an essay on therapeutic cancer vaccines. He compared early attempts to target proteins shared across tumors with personalized mRNA therapies made for individual patients.

For many years, researchers searched for proteins found in tumors from different people. MAGE-A3 is one such target. It was evaluated in phase III trials in melanoma and non-small-cell lung cancer. In both trials, treatment with MAGE-A3 failed to improve disease-free survival.

A T-cell is an immune cell that can kill a tumor cell. It recognizes a short protein fragment, called a peptide, on the tumor cell's surface. An MHC molecule holds the peptide and presents the cell's contents to the immune system. To design a personalized vaccine, researchers compare DNA from the patient's tumor with DNA from the patient's normal tissue. Tumor mutations provide candidate neoantigens, which are altered protein fragments produced in the tumor. Researchers then select peptides that MHC molecules can present to T-cells and that can trigger a T-cell response.

“Cancer vaccines are not a drug in the conventional sense. They are a manufacturing process,” Mahajan writes.

Sequencing reads the DNA sequence and produces a list of mutations. mRNA is a molecule that provides instructions to cells and carries information about the selected neoantigens. Pembrolizumab blocks PD-1, a protein that tumors can use to suppress T-cells.

On August 19, Merck and Moderna reported results from a prespecified interim analysis of the phase III INTerpath-001 trial. The combination of the personalized mRNA therapy intismeran autogene and pembrolizumab met both stated endpoints, recurrence-free survival and distant metastasis-free survival, in people with completely resected stage IIB–IV cutaneous melanoma. The control group received pembrolizumab with placebo, so the comparison measures the contribution of the personalized therapy to the postoperative regimen.

Mahajan connects the history of MAGE-A3 with this trial through a single practical task. For each patient, researchers must identify tumor mutations, select fragments that the immune system can recognize, manufacture the therapy, and test it together with postoperative immunotherapy.